A research menu p.1More spending on agricultural science is needed to help resolve the world's food crisis.
News & Views
Hidden female talent p.41A male fruitfly serenades his female with a courtship song produced by vibrating one wing. The female also has the neuronal circuitry to generate a song of her own, but her brain tells her not to.
The fourth element p.42Almost four decades since its existence was first proposed, a fourth basic circuit element joins the canonical three. The 'memristor' might herald a step-change in the march towards ever more powerful circuitry.
The squeeze goes on p.45After 20 years of hard labour, squeezed states — light and matter whose quantum fluctuations have been arduously suppressed below standard levels of quantum noise — are coming of age and are ripe for application.
Rays from the dark p.48The origin of the cosmic rays that bombard Earth has troubled physicists for nigh on a century. Supernova remnants are a favoured source — but we should keep our minds open to alternatives.
All metazoan cells carry transmembrane receptors of the integrin family, which couple the contractile force of the actomyosin cytoskeleton to the extracellular environment. In agreement with this principle, rapidly migrating leukocytes use integrin-mediated adhesion when moving over two-dimensional surfaces. As migration on two-dimensional substrates naturally overemphasizes the role of adhesion, the contribution of integrins during three-dimensional movement of leukocytes within tissues has remained controversial. We studied the interplay between adhesive, contractile and protrusive forces during interstitial leukocyte chemotaxis in vivo and in vitro. We ablated all integrin heterodimers from murine leukocytes, and show here that functional integrins do not contribute to migration in three-dimensional environments. Instead, these cells migrate by the sole force of actin-network expansion, which promotes protrusive flowing of the leading edge. Myosin II-dependent contraction is only required on passage through narrow gaps, where a squeezing contraction of the trailing edge propels the rigid nucleus.
Genetic variation among individual humans occurs on many different scales, ranging from gross alterations in the human karyotype to single nucleotide changes. Here we explore variation on an intermediate scale—particularly insertions, deletions and inversions affecting from a few thousand to a few million base pairs. We employed a clone-based method to interrogate this intermediate structural variation in eight individuals of diverse geographic ancestry. Our analysis provides a comprehensive overview of the normal pattern of structural variation present in these genomes, refining the location of 1,695 structural variants. We find that 50% were seen in more than one individual and that nearly half lay outside regions of the genome previously described as structurally variant. We discover 525 new insertion sequences that are not present in the human reference genome and show that many of these are variable in copy number between individuals. Complete sequencing of 261 structural variants reveals considerable locus complexity and provides insights into the different mutational processes that have shaped the human genome. These data provide the first high-resolution sequence map of human structural variation—a standard for genotyping platforms and a prelude to future individual genome sequencing projects.
Regulatory T cells (Treg) expressing the transcription factor Foxp3 control the autoreactive components of the immune system. The development of Treg cells is reciprocally related to that of pro-inflammatory T cells producing interleukin-17 (TH17). Although Treg cell dysfunction and/or TH17 cell dysregulation are thought to contribute to the development of autoimmune disorders, little is known about the physiological pathways that control the generation of these cell lineages. Here we report the identification of the ligand-activated transcription factor aryl hydrocarbon receptor (AHR) as a regulator of Treg and TH17 cell differentiation in mice. AHR activation by its ligand 2,3,7,8-tetrachlorodibenzo-p-dioxin induced functional Treg cells that suppressed experimental autoimmune encephalomyelitis. On the other hand, AHR activation by 6-formylindolo[3,2-b]carbazole interfered with Treg cell development, boosted TH17 cell differentiation and increased the severity of experimental autoimmune encephalomyelitis in mice. Thus, AHR regulates both Treg and TH17 cell differentiation in a ligand-specific fashion, constituting a unique target for therapeutic immunomodulation.
Dust near Jupiter is produced when interplanetary impactors collide energetically with small inner moons, and is organized into a main ring, an inner halo, and two fainter and more distant gossamer rings
Jupiter’s dusty rings were discovered by the Voyager spacecraft in 1979 and have subsequently been investigated with Earth- and space-based telescopes
A unique source of data on the outer jovian rings was provided by Galileo, which passed through the tenuous Thebe and Amalthea rings in 2002 and 2003 (
The orbits of micrometre-sized ring particles are dominated by Jupiter’s gravity, but are also subject to electromagnetic forces and, to a far lesser extent, radiation pressure from the Sun. Dust grains acquire electric charges in Jupiter’s vicinity by sweeping up local plasma and by photoemission
Without planetary shadowing, the effects of electromagnetic forces on the sizes (semimajor axes a) and shapes (eccentricities e) of ring particle orbits average to zero over satellite orbital periods (∼12 h) (ref.
As the shadow resonance is a fully electromagnetic effect and magnetic fields do no work, energy measured in the reference frame that co-rotates with Jupiter’s field is conserved. This constraint leads to a constant of the motion (the dimensionless Jacobi constant):
Particles blasted off satellites start on nearly circular orbits around Jupiter, because satellite orbital speeds (∼25 km s
We ran a number of dynamical simulations including gravitational, electromagnetic and radiation forces
We now compare this suite of models to the other observations discussed above. The visual appearance of the Thebe extension in
There are several reasons why the Amalthea ring does not show a similar extension
Galileo’s dust detector observed a significant population of particles with orbits tilted by ≳20°. The models indicate that for certain plasma conditions, particle orbits do in fact attain large inclinations (
Finally, Galileo observed a strong excess of 0.2–0.3 &mgr;m dust grains at Amalthea (
We find that the shadow resonance at Jupiter can simultaneously account for (1) the outward extension of the Thebe ring at constant inclination, (2) the gap interior to Thebe, (3) the large number of high-inclination particles, and (4) the excess of 0.2–0.3 &mgr;m grains at Amalthea in a natural model with only a single tunable parameter: the plasma density, which we constrain to 0.5 cm
a, Jupiter’s gossamer rings as seen by Galileo from a location near the planet’s equatorial plane (false colour: adapted from ref.
a, Impact rate of submicrometre-sized dust particles onto the Galileo dust detector
a, Electric potential for 3.2 &mgr;m dust grains from Amalthea (dotted lines) and Thebe (solid lines) as a function of time. We show charge histories for three different assumed plasma densities ne in a simple two-component plasma model as dust grains travel from sunlight (flat curves) into Jupiter’s shadow (dips to lower potentials) and out again. Dust in a plasma tends to charge negatively, and higher plasma densities lead to lower grain potentials, both in sunlight and in shadow. Sunlight raises equilibrium potentials because the photoelectric effect ejects electrons from a dust grain’s surface. In reality, plasma properties probably vary with distance both from Jupiter and the equatorial plane, but we find that these non-uniform distributions lead to changes in orbital behaviour only over timescales much longer than those considered here, validating our simplifying assumption (compare ref.
Here we plot numerically determined minimum and maximum distances (left plots) and latitudes (right plots) reached by different-sized ring particles within two years of launch from Thebe (solid curves) and Amalthea (dotted curves). Three different charging histories parameterized by ne and 600 separate numerical simulations are presented here. The mobility of dust grains is enhanced by two factors: (1) small particle sizes which have larger electromagnetic perturbations, and (2) slow pericentre precession rates which allow the shadow resonance to pump eccentricities to extremely high values
We thank the Galileo project for a successful mission. This research was supported by grants from NASA Exobiology (D.P.H.) and DLR (H.K.).
Author Contributions H.K. did all of the data analysis and created
Phase-slips control dissipation in many bosonic systems, determining the critical velocity of superfluid helium
Although it is widely believed to be the simplest model that captures the relevant features of boson physics in a variety of physical systems, the Bose–Hubbard model is not integrable, and a full characterization of its features is therefore a challenging theoretical problem. The Bose–Hubbard model is described by the hamiltonian
We study mass transport, which is the equivalent of charge transport for neutral bosons. Charge transport is studied in solids by using an electric field to apply a uniform force to the charge carriers. In contrast, we use the spatially inhomogeneous restoring force from a parabolic confining potential to excite damped harmonic motion of the centre of mass (COM) of the gas of atoms. The confining potential is included in
Previous work on the transport of Bose–Einstein condensates (BECs) in optical lattices has focused on regimes not described by the Bose–Hubbard model
The experimental sequence is shown in
The COM motion we observe is described well by damped harmonic motion;
The temperature dependence of the damping rate &ggr; is shown in
The damping rate data are plotted against inverse temperature because we fit the data to a model,
The data shown in
Other dissipation mechanisms may contribute to the damping rate at high temperature. For example, mutual friction between the BEC and the thermal component may play a role at high temperature and for large values of s, where we observe relative motion between the BEC and the thermal component (see
In the quantum tunnelling regime, the phase-slip rate is predicted to be proportional to e
In conclusion, we observe temperature-dependent damping of COM motion in an optical lattice that is consistent with dissipation caused by phase slips. The parabolic potential used to confine the atoms gives rise to an inhomogeneous density distribution, which may enhance the effect of phase slips because the activation barrier is suppressed at the edge of the BEC. This system may therefore be comparable to thin superconducting wires and strips, in which vortices entering and leaving at the boundaries strongly influence current flow
The optical lattice is created using three pairs of orthogonally polarized laser beams with wavelengths of 812 nm. These beams are weakly focused to a 120-&mgr;m waist, and slightly frequency-offset to eliminate residual cross-dimensional interference resulting from imperfect polarization. COM motion is excited by changing the strength of a confining magnetic potential for 5 ms. The signal-to-noise ratio in measurements of the COM velocity is improved by reducing the lattice potential to zero over an interval of 200 &mgr;s, thereby eliminating atoms effectively diffracting from the lattice potential.
The temperature is determined by measuring the fraction of atoms in the BEC. For our lowest temperature data (T/Tc = 0.6 and T/Tc = 0.3), accurately detecting the atoms in the thermal component requires images taken at high and low optical depths.
The oscillation data used to measure &ggr; are checked in two ways for nonlinear response. First, no significant change in the fitted &ggr; value is measured if the first period of motion is excluded from the fit. Second, fitting the data to a nonlinear damping model with
Standard absorption imaging is used to measure the COM position of the atoms, which is translated into COM velocity, after release from the optical lattice and the harmonic potential. Before turning off the magnetic trap, the lattice is linearly ramped down over a 200-&mgr;s interval to improve the image signal-to-noise ratio. We checked that this ‘band-mapping’
We determine temperature for T/Tc > 0.6 by measuring the fraction of atoms in the BEC using a single absorption image taken after the BEC’s release from the magnetic trap. The atoms are optically pumped (or ‘repumped’) to the
We measure T/Tc using the fraction of atoms in the BEC because there are no proven techniques for determining T in an optical lattice. Interactions may change T/Tc for BECs loaded into the lattice for the highest value of s in
We rule out several technical noise sources as dissipation mechanisms that could explain the damping of COM motion. Anharmonicity in the dipole potential may effectively damp COM motion for large values of s. To check for anharmonic behaviour, we measure COM motion when the retro-reflected lattice laser beams are removed, which eliminates the lattice potential and reduces the depth of the dipole potential by a factor of ∼2. Using this technique we measure a value of &ggr; consistent with zero for lattice laser intensities corresponding to s = 9 (or s = 18 if the retro-reflected beams are present), eliminating trap anharmonicity as an effective dissipation source for the data in
Over a 100-ms interval a
Each data point represents the result of one experimental cycle, in which the COM velocities of the BEC (filled circles) and the thermal component (open circles) are measured. a, Data for s = 2, T/Tc = 0.85 and 8 × 10
a, s = 6; b, s = 2. Each data point represents the value of &ggr; determined from BEC COM motion at fixed temperature and fixed s. The red data points are excluded from the fit (solid line) to a model of thermally activated damping. The fitted energy gap &Dgr;E is kB × (0.5 ± 0.1) &mgr;K for s = 2 and kB × (0.33 ± 0.08) &mgr;K for s = 6 (where kB is Boltzmann’s constant). The error bars in &ggr; represent the uncertainty in the fit to COM velocity data. The standard deviations in the data used to determine T/Tc for each data point are used for the error bars in (T/Tc)
a, The tight-binding limit is used to calculate
We thank R. Barankov, E. Demler, P. Goldbart, N. Goldenfeld, D. Pekker and P. Phillips for discussions, and D. S. Jin, N. Mason and J. V. Porto for critically reading this manuscript. This work was supported by the National Science Foundation, the Office of Naval Research and the UIUC Research Board. D.M. acknowledges support from the Carver Foundation and NSERC.
Anyone who ever took an electronics laboratory class will be familiar with the fundamental passive circuit elements: the resistor, the capacitor and the inductor. However, in 1971 Leon Chua reasoned from symmetry arguments that there should be a fourth fundamental element, which he called a memristor (short for memory resistor)
More specifically, Chua noted that there are six different mathematical relations connecting pairs of the four fundamental circuit variables: electric current i, voltage v, charge q and magnetic flux ϕ. One of these relations (the charge is the time integral of the current) is determined from the definitions of two of the variables, and another (the flux is the time integral of the electromotive force, or voltage) is determined from Faraday’s law of induction. Thus, there should be four basic circuit elements described by the remaining relations between the variables (
In the case of linear elements, in which M is a constant, memristance is identical to resistance and, thus, is of no special interest. However, if M is itself a function of q, yielding a nonlinear circuit element, then the situation is more interesting. The i–v characteristic of such a nonlinear relation between q and ϕ for a sinusoidal input is generally a frequency-dependent Lissajous figure
The most basic mathematical definition of a current-controlled memristor for circuit analysis is the differential form
Here we present a physical model of a two-terminal electrical device that behaves like a perfect memristor for a certain restricted range of the state variable w and as a memristive system for another, wider (but still bounded), range of w. This intuitive model produces rich hysteretic behaviour controlled by the intrinsic nonlinearity of M and the boundary conditions on the state variable w. The results provide a simplified explanation for reports of current–voltage anomalies, including switching and hysteretic conductance, multiple conductance states and apparent negative differential resistance, especially in thin-film, two-terminal nanoscale devices, that have been appearing in the literature for nearly 50 years
Electrical switching in thin-film devices has recently attracted renewed attention, because such a technology may enable functional scaling of logic and memory circuits well beyond the limits of complementary metal–oxide–semiconductors
The application of an external bias v(t) across the device will move the boundary between the two regions by causing the charged dopants to drift
The coupled equations of motion for the charged dopants and the electrons in this system take the normal form for a current-controlled (or charge-controlled) memristor (
In nanoscale devices, small voltages can yield enormous electric fields, which in turn can produce significant nonlinearities in ionic transport.
The model of
The rich hysteretic i–v characteristics detected in many thin-film, two-terminal devices can now be understood as memristive behaviour defined by coupled equations of motion: some for (ionized) atomic degrees of freedom that define the internal state of the device, and others for the electronic transport. This behaviour is increasingly relevant as the active region in many electronic devices continues to shrink to a width of only a few nanometres, so even a low applied voltage corresponds to a large electric field that can cause charged species to move. Such dopant or impurity motion through the active region can produce dramatic changes in the device resistance. Including memristors and memristive systems in integrated circuits has the potential to significantly extend circuit functionality as long as the dynamical nature of such devices is understood and properly used. Important applications include ultradense, semi-non-volatile memories and learning networks that require a synapse-like function.
Resistors and memristors are subsets of a more general class of dynamical devices, memristive systems. Note that R, C, L and M can be functions of the independent variable in their defining equations, yielding nonlinear elements. For example, a charge-controlled memristor is defined by a single-valued function M(q).
a, Diagram with a simplified equivalent circuit. V, voltmeter; A, ammeter. b, c, The applied voltage (blue) and resulting current (green) as a function of time t for a typical memristor. In b the applied voltage is v0sin(&ohgr;0t) and the resistance ratio is
a, Simulation with dynamic negative differential resistance; b, simulation with no dynamic negative differential resistance; c, simulation governed by nonlinear ionic drift. In the upper plots of a, b and c we plot the voltage stimulus (blue) and the corresponding change in the normalized state variable w/D (red), versus time. In all cases, hard switching occurs when w/D closely approaches the boundaries at zero and one (dashed), and the qualitatively different i–v hysteresis shapes are due to the specific dependence of w/D on the electric field near the boundaries. d, For comparison, we present an experimental i–v plot of a Pt–TiO2-x–Pt device
This research was conducted with partial support from DARPA and DTO.
The file contains Supplementary Discussion on global correlation skill and root mean square error maps; surface temperature maps for 1994-2004 and 2005-2015 decades; and Supplementary Figures 1-4 with Legends.
The climate of the North Atlantic region exhibits fluctuations on decadal timescales that have large societal consequences. Prominent examples include hurricane activity in the Atlantic
The North Atlantic is a region with large natural multidecadal variability
Studies with coupled general circulation models assuming perfect ocean initial conditions indicate that accurate initialization of MOC may allow Atlantic multidecadal variability to be predicted a decade or more in advance
Using this simple initialization technique and a state-of-the-art coupled general circulation model
Skilful predictions of surface temperature averaged over years one to ten of the hindcasts are obtained over large parts of the North Atlantic, Europe, North America and northern Africa (
In several regions, initialization causes a significant degradation in skill compared to the twentieth century-RF simulations (
Skill in predicting North Atlantic SST on decadal timescales would imply, as described above, skill in initializing the Atlantic MOC. In the SST-restored simulations, the Atlantic MOC weakens from the 1950s till the mid-1960s, strengthens thereafter, peaking between the late-1980s to late-1990s, and subsequently weakens again (
In contrast to measurements of the MOC, regular hydrographic observations in the Labrador Sea extend back past the 1950s. As density changes in the Labrador Sea are widely accepted to force MOC variations
On multidecadal timescales, Labrador Sea convection is initialized directly by SST relaxation. The latter includes the history of observed atmospheric forcing, and in particular the North Atlantic Oscillation (NAO), which plays a key role in forcing multidecadal MOC variations
Initialized decadal fluctuations in the Atlantic MOC are predictable a decade in advance, with ensemble-spread small compared to the signal (
Enhanced skill over the twentieth century-RF simulations in predicting land surface temperature (
We now consider two forecasts, started in November 2000 and November 2005. The MOC is predicted to weaken almost to its 1950–2005 mean over the next decade (
To investigate the sensitivity of the predictions to greenhouse gas forcing only, the two forecasts were repeated assuming that greenhouse gases were stabilized at year 2000 values. The predictions for the MOC and surface temperature remain basically unchanged. Thus, in the near future, natural decadal variability in the Atlantic and Pacific may not only override the regional effects of global warming, but temporarily weaken it. Thus, a joint initial/boundary value problem has to be considered when forecasting North Atlantic sector and global climate variability for the coming decades.
The results presented here are promising in light of existing model biases. Experience in numerical weather prediction and seasonal forecasting has shown that skill can be considerably improved by reducing model systematic error and by more accurate forecast initialization. Thus, useful decadal predictions may be in reach.
Results are based on three sets of simulations with the ECHAM5/MPI-OM coupled general circulation model
For nine forecasts we assume seven degrees of freedom (d.f.). As only positive correlations indicate skill, the significance level for the forecast is determined using a one-sided t-test. A correlation is determined to be significantly larger or smaller than another if the Fisher-Z transformed values pass a one-sided t-test for differences in means (with 12 d.f.). The Fisher-Z distribution of correlations of nine independent and normally distributed pairs of points is close to normal. Field significance is estimated on a 15° × 15° grid to minimize effects of spatial correlation.
a, Skill of nine ten-year-long predictions, evenly distributed over the period 1955–2005, made with a climate model initialized using ocean (SST) observations and run with projected changes in radiative forcing. b, As in a but given by persistence. c, As in a, but not initialized using ocean observations and with radiative forcing following observations. d, As in c, but with model SST relaxed to observations between 60° S and 60° N (seen in near perfect correlations over the ocean). Correlations exceeding 0.58 are significant at the 5% level. Regions where initialization results in a significant enhancement or reduction in skill compared to radiative-forcing-only simulations are indicated by a blue cross-hatching in a and c, respectively. Land regions where restoring to observed SST anomalies provides a significant enhancement in skill relative to radiative-forcing-only simulations are indicated by blue cross-hatching in d. Correlations in a and c are field significant at close to the 0% level, while those in b pass the field significance test at the 1% level. Details of significance estimation are given in Methods. Correlations in a, c and d are computed from the ensemble mean of three simulations. SST observations are from HADISST
a, Annual mean meridional overturning circulation (MOC) at 30° N from the SST-restored simulations. Grey shading indicates ensemble spread in all panels. b, Simulated wintertime (December–February) Labrador Sea (60–50° W, 55–65° N) mixed-layer depth and observed annual mean Labrador Sea Water (LSW) thickness
Model projections are twentieth century-RF followed by A1B scenario simulations (‘20C-RF/A1B’). a, Maximum MOC strength at 30° N; b, Atlantic SST dipole index (60–10° W, 40–60° N minus 50–0° W, 40–60° S SST area averages), which is constructed to isolate MOC forced SST fluctuations from radiatively forced variations
Model projections are twentieth century-RF followed by A1B scenario simulations (‘20C-RF/A1B’); ‘Stabilization’ forecasts assume greenhouse gas concentrations fixed at year 2000 levels. Each point represents a ten-year centred mean; vertical bars indicate ensemble spread; verification and forecast periods are indicated (dark shading begins 2008, indicating the start of the true forecast period). Three additional decadal means (joined by a dotted line) show the evolution of the initialized and un-initialized 2005 predictions extended till 2030. Correlation of both hindcasts and climate model projections with observations are given in brackets. Correlation of the twentieth century-RF simulation with observations is greater than that of the hindcasts, but only marginally at the 5% significance level. Observed global mean temperature anomalies are from HadCRU3
Computing resources at the European Centre for Medium Range Weather Forecasting, the Deutsches Klimarechenzentrum and the Norddeutscher Verbund für Hoch- und Höchstleistungsrechnen are acknowledged. Deutscher Wetterdienst provided significant CPU allocations. The work was supported by the European Union ENSEMBLES and German BMBF NORDATLANTIK projects. We thank M. Esch for performing the three twentieth century-RF simulations, and F. Nevoigt for help with art work.
Author Contributions N.S.K. and M.L. designed experiments. N.S.K. performed experiments and analysis. M.L. and J.J. contributed to analysis. L.K., J.J. and E.R. provided the model and support. N.S.K. and M.L. wrote the paper. All authors discussed the results and commented on the manuscript.
The file contains Supplementary Tables S1-S3, Supplementary Discussion, Supplementary Figures S1-S4 with legends and additional references.
The Gakkel ridge, which extends under the Arctic ice cap for ∼1,800 km, is the slowest spreading ocean ridge on Earth. Its spreading created the Eurasian basin, which is isolated from the rest of the oceanic mantle by North America, Eurasia and the Lomonosov ridge. The Gakkel ridge thus provides unique opportunities to investigate the composition of the sub-Arctic mantle and mantle heterogeneity and melting at the lower limits of seafloor spreading. The first results of the 2001 Arctic Mid-Ocean Ridge Expedition (ref.
The Gakkel ridge is the global ultraslow endmember among ocean ridges, with spreading rates decreasing from ∼15 mm yr
Prior to the 2001 Arctic Mid-Ocean Ridge Expedition, data on Gakkel lavas were restricted to three samples of basaltic cobbles and glass shards recovered on two cruises
Our data confirm fundamental differences in the upper-mantle sources of WVZ and EMZ lavas (locations and data tables can be found in the
The contrasting WVZ and EVZ mantle compositions are shown in
Gakkel WVZ lavas, however, differ from Southern Hemisphere Dupal basalts, in that they have low &Dgr;7/4 values
A striking feature of the Gakkel ridge is the abruptness of the isotopic boundary and its location in the middle of the SMZ (
Comparison with global MORB (
The small geographical scale and well-understood tectonic history of the Eurasian basin affords opportunities to explore how the Dupal signature is created. Quaternary alkaline basalts ranging from hawaiites to nephelinitic basanites in northwestern Spitsbergen have been attributed to low-degree melting of subcontinental lithospheric mantle (SCLM)
Insight into the mineral assemblage of the Spitsbergen magma source can be gleaned from the comparison of elements more compatible in amphibole or phlogopite (such as Rb, Ba and K) with other elements unaffected by amphibole or phlogopite (such as Th, U and La). In plots of
The data offer insight into the long-term history of the Gakkel asthenosphere. A Rb–Sr ‘pseudo-isochron’ plot (
The introduction of the Svalbard SCLM to the WVZ asthenosphere, the leakage of North Atlantic mantle into the Arctic, and the formation of the boundary can be understood in the context of Arctic tectonic history (
The origin of the Southern Hemisphere Dupal signature is an ongoing debate in mantle dynamics, and a recent summary highlights inadequacies of postulated models
In summary, the Gakkel ridge is an ultraslow-spreading ridge surrounded by continental lithosphere, with clear evidence of regional SCLM contributing a Dupal geochemical signature to the asthenospheric mantle. Unlike in other regions on Earth, we have the strong evidence of Spitsbergen volcanics, with the requisite isotopic compositions to be the Gakkel endmember, and a source mineralogy definitively indicating SCLM melting. The process documented in the Arctic is chemically and geologically feasible and probably plays an important role in generating chemical heterogeneity along the mid-ocean ridge system. Dupal signatures globally may have multiple sources; nevertheless, the migrating continents that have traversed the Indian Ocean basin, where Dupal characteristics occur on a large scale, would generate substantial opportunities for analogous SCLM contributions to the sub-Indian Ocean mantle.
All analyses were made on hand-picked basaltic glasses. Major elements were measured on a Cameca electron microprobe at the American Museum of Natural History, operated as a joint facility with the Lamont-Doherty Earth Observatory of Columbia University (LDEO). Trace elements were measured using a Plasma Quad 2 inductively coupled plasma mass spectrometer at LDEO or a Thermo Electron X-series inductively coupled plasma mass spectrometer at Harvard. Pb, Nd and Sr isotope ratios were measured using a VG Sector 54 multicollector thermal ionization mass spectrometer at LDEO.
Major elements were measured at the American Museum of Natural History on the AMNH-LDEO Cameca electron microprobe. Each analysis represents an average of five individual measurements on a single glass chip. A glass chip of the Lamont glass probe standard, JDF-D2, was repeatedly analysed in each probe session to monitor long-term drift. No long-term drift was evident in any run. The errors are based on the reproducibility of JDF analyses over multiple probe sessions and are ≤1% for SiO2 and MgO.
Hand-picked glass separates with minimal surface alteration were analysed for trace element compositions and measured on the Plasma Quad 2 inductively coupled plasma mass spectrometer at LDEO or a Thermo Electron X-series inductively coupled plasma mass spectrometer at Harvard . Solutions were spiked with Ge, In, Tm and Bi to correct for in-run drift. Concentrations were determined using a calibration curve based on US Geological Survey and in-house standards measured during each run. Errors are 3–5%.
Pb, Nd and Sr isotope ratios were measured on separate glass chips from the trace element analyses. These were leached cold for 12 minutes with 8N HNO3. Pb was separated using AG1-X8 anion resin, Sr was separated using Eichrom Sr resin and Nd was separated in a two-column procedure using Eichrom TRU-spec resin to separate the rare-earth elements, followed by &agr;-hydroxy isobutyric acid. Sr, Nd and Pb isotopes were measured on a VG Sector 54 multicollector thermal ionization mass spectrometer at LDEO.
a, Approximately 53 Myr ago, the nascent Eurasian basin, underlain by ‘Arctic’ upper mantle (labelled ‘A’) as sampled today in the EVZ, was separated from the North Atlantic ridges (labelled ‘NA’) by a continent. b, During the Cenozoic, the Eurasian basin continued to widen, and the North Atlantic ridges propagated north, with Greenland and Svalbard separating ∼10 Myr ago. c, Present-day situation based on the Gakkel data, with the boundary between the North Atlantic and Arctic upper-mantle provinces well within the Eurasian basin, in the middle of the SMZ. ‘YP’ labels the Yermak plateau, near the WVZ. Images modified from Blythe and Kleinspehn
We thank A. le Roex for comments that helped to improve this paper. This work was supported by the US National Science Foundation. G.S. was supported partly by a Paul and Daisy Soros Fellowship for New Americans.
The geochemical data reported here are available in the Petrological Database of the Ocean Floor (
The file contains Supplementary Figure with Legend showing longitudinal section of recently fertilized ovule of Hydatella inconspicua.
The flowering plant family Hydatellaceae was recently discovered to be allied to the ancient angiosperm lineage Nymphaeales (water lilies)
For over a century, the flowering plant family Hydatellaceae was thought to belong to the Poales, a highly derived monocot order that includes the grasses
Although the data are, at best, ambiguous, the original embryological studies of Hydatellaceae
Working with field-collected material of Hydatella inconspicua, I examined female gametophyte development and ontogenetic features of the formation of embryo-nourishing reserves in the ovule/seed. The mature female gametophyte of H. inconspicua contains four uninucleate cells: the egg cell, two synergids and a uninucleate central cell (
Critically, at the two-nucleate syncytial stage in H. inconspicua, both nuclei are located at the micropylar pole of the female gametophyte (
Although endosperm is initiated in H. inconspicua (
Remarkably, the embryo-nourishing tissue within the ovule/seed in Hydatella begins to acquire significant carbon resources (starch) from the maternal plant before fertilization. At the two-nucleate stage of pre-fertilization female gametophyte development, the perisperm contains reserves of starch (
Angiosperms (with the exception of Hydatellaceae) differ from extant gymnosperms (conifers, Ginkgo, cycads, Gnetales) in that maternal commitment of embryo-nourishing resources to seeds only occurs after fertilization
The exceptional case of pre-fertilization maternal provisioning of a seed in Hydatella may well be an apomorphic feature of Hydatellaceae alone. If so, Hydatellaceae stands as an essentially unique angiosperm clade that has reverted to allocating resources to an ovule/seed before the initiation of an embryo (albeit to the perisperm and not to the female gametophyte, as in gymnosperms). Because most taxa of flowering plants use a sexually formed endosperm to nourish an embryo, pre-fertilization allocation of embryo-nourishing reserves to the seed is ontogenetically precluded. However, members of a large number of disparate clades of angiosperms, including many basal lineages (Nymphaeaceae, Cabombaceae, Trimeniaceae, Acoraceae, Ceratophyllaceae, Saururaceae, Piperaceae, Hydnoraceae), form an embryo-nourishing perisperm
Alternatively, pre-fertilization maternal resource allocation to ovules/seeds, and specifically to a maternally derived perisperm, in Hydatella could represent a plesiomorphic and transitional condition associated with the origin of flowering plants. If so, the ‘underdeveloped’ endosperm of Hydatellaceae and Nymphaeales is not reduced, but rather represents an intermediate condition (between gymnosperms and other angiosperms) in which the endosperm has not yet achieved its fully fledged role as the primary source of nutrients for the developing embryo. It is important to note that Amborella and most members of the Austrobaileyales contain a well-developed endosperm
Charles Darwin was among the first to recognize the immense chasm between gymnosperms and angiosperms with respect to their biological characteristics (letters to Oswald Heer in 1875 and to Joseph Hooker in 1879 and 1882). Since then, inferring the vegetative, floral and reproductive features that defined the first angiosperms, as well as their evolutionary (transformational) links to a gymnospermous ancestor, has proved to be fraught with difficulties. Indeed, the past five years have witnessed the near-global collapse of a century-old set of paradigms concerning the embryological features of the earliest angiosperms
Hydatella inconspicua was collected on 6 December 2006 at Kai Iwi Lake, Northland, New Zealand, by P. Champion, New Zealand National Institute of Water and Atmospheric Research.
Plants were chemically fixed in 4% glutaraldehyde, washed in phosphate buffer and stored in water. Specimens were dehydrated through an ethanol series, then infiltrated and embedded in glycol methacrylate (
The micropylar pole is towards the top of the figure. At the two-nucleate syncytial stage (blue boxes), water-lily female gametophytes have both nuclei at the micropylar pole; in Polygonum-type female gametophytes, a nuclear migration event leads to placement of a single nucleus at each pole. Red, synergids; yellow, egg cell; brown, antipodals. cc, central cell. pn, polar nucleus.
a, Two-nucleate syncytial female gametophyte in H. inconspicua, with both nuclei (red arrowheads) at micropylar pole. Grey box contains digital superposition of second nucleus from adjacent histological section. b, Mature four-celled female gametophyte in H. inconspicua. Grey box contains digital superposition of second synergid nucleus from adjacent histological section. cc, central cell; ec, egg cell; pn, polar nucleus; sc, synergid cell.
a, Longitudinal section of perisperm in pre-fertilization ovule at two-nucleate stage of female gametophyte development. b, Cross-polarization optical image demonstrating presence of starch from boxed region in a. Starch grains, which are birefringent, appear as four white quadrants separated by a black cross. c, Longitudinal section of perisperm in pre-fertilization ovule just before fertilization. d, Cross-polarization optical image demonstrating presence of large amounts of starch from boxed region in c. Starch grains, which are birefringent, appear as four white quadrants separated by a black cross. e, Longitudinal section of perisperm in post-fertilization seed with three-celled embryo and minimally developed endosperm. f, Higher magnification from boxed region in e of embryo surrounded by a small endosperm tissue that lacks significant storage reserves. The perisperm below the endosperm contains large amounts of starch (bright circular structures). Scale bars, 10 &mgr;m. em, embryo; es, endosperm; fg, female gametophyte; ps, perisperm.
I thank: P. Champion and A. Drinnan for collecting plant materials; S. Holloway for histological work; S. Renner for translation of the embryological studies of U. Hamaan; and P. Diggle, L. Hufford, J. Williams and R. Robichaux for feedback on this manuscript. This work was supported by a National Science Foundation Research Grant.
This file contains Supplementary Notes including the technical details of our hierarchical model and the methods used to fit it to empirical data. It also contains addition results on graph resampling and the prediction of missing links, and the algorithmic specifics of our experimental studies.
Networks have in recent years emerged as an invaluable tool for describing and quantifying complex systems in many branches of science
Much recent work has been devoted to the study of clustering and community structure in networks
This model, which we call a hierarchical random graph, is similar in spirit to (although different in realization from) the tree-based models used in some studies of network search and navigation
Given a dendrogram and a set of probabilities pr, the hierarchical random graph model allows us to generate artificial networks with a specified hierarchical structure, a procedure that might be useful in certain situations. Our goal here, however, is a different one. We wish to detect and analyse the hierarchical structure, if any, of networks in the real world. We accomplish this by fitting the hierarchical model to observed network data by using the tools of statistical inference, combining a maximum-likelihood approach
The success of this approach relies on the flexible nature of our hierarchical model, which allows us to fit a wide range of network structures. The traditional picture of communities or modules in a network, for example, corresponds to connections that are dense within groups of vertices and sparse between them—a behaviour called ‘assortativity’ in the literature
To demonstrate our method we have used it to construct hierarchical decompositions of three example networks drawn from disparate fields: the metabolic network of the spirochaete Treponema pallidum
The dendrograms produced by our method are also of interest in themselves, as a graphical representation and summary of the hierarchical structure of the observed network. As discussed above, our method can generate not just a single dendrogram but a set of dendrograms, each of which is a good fit to the data. From this set we can, by using techniques from phylogeny reconstruction
Another application of the hierarchical decomposition is the prediction of missing interactions in networks. In many settings, the discovery of interactions in a network requires significant experimental effort in the laboratory or the field. As a result, our current pictures of many networks are substantially incomplete
The hierarchical decomposition can be used as the basis for an effective method of predicting missing interactions as follows. Given an observed but incomplete network, we generate, as described above, a set of hierarchical random graphs—dendrograms and the associated probabilities pr—that fit that network. Then we look for pairs of vertices that have a high average probability of connection within these hierarchical random graphs but are unconnected in the observed network. These pairs we consider the most likely candidates for missing connections. (Technical details of the procedure are given in
We demonstrate the method by using our three example networks again. For each network we remove a subset of connections chosen uniformly at random and then attempt to predict, on the basis of the remaining connections, which have been removed. A standard metric for quantifying the accuracy of prediction algorithms, commonly used in the medical and machine learning communities, is the AUC statistic, which is equivalent to the area under the receiver operating characteristic (ROC) curve
However, these simple methods can be misleading for networks that exhibit more general types of structure. In food webs, for instance, pairs of predators often share prey species but rarely prey on each other. In such situations a common-neighbour or shortest-path-based method would predict connections between predators where none exists. The hierarchical model, by contrast, is capable of expressing both assortative and disassortative structure and, as
In the applications above, we have assumed for simplicity that there are no false positives in our network data; that is, that every observed edge corresponds to a real interaction. In networks in which false positives may be present, however, they too could be predicted by using the same approach: we would simply look for pairs of vertices that have a low average probability of connection within the hierarchical random graph but are connected in the observed network.
The method described here could also be extended to incorporate domain-specific information, such as species’ morphological or behavioural traits for food webs
In closing, we note that our approach differs crucially from previous work on hierarchical structure in networks
Computer code implementing many of the analysis methods described in this paper can be found online at
Each internal node r of the dendrogram is associated with a probability pr that a pair of vertices in the left and right subtrees of that node are connected. (The shades of the internal nodes in the figure represent the probabilities.)
a, Consensus dendrogram reconstructed from the sampled hierarchical models. b, A visualization of the network in which the upper few levels of the consensus dendrogram are shown as boxes around species (plants, herbivores, parasitoids, hyperparasitoids and hyper-hyperparasitoids are shown as circles, boxes, down triangles, up triangles and diamonds, respectively). Note that in several cases a set of parasitoids is grouped into a disassortative community by the algorithm, not because they prey on each other but because they prey on the same herbivore.
Average AUC statistic—that is, the probability of ranking a true positive over a true negative—as a function of the fraction of connections known to the algorithm, for the link prediction method presented here and a variety of previously published methods. a, Terrorist association network; b, T. pallidum metabolic network; c, grassland species network.
We thank J. Dunne, M. Gastner, P. Holme, M. Huss, M. Porter, C. Shalizi and C. Wiggins for their help, and the Santa Fe Institute for its support. C.M. thanks the Center for the Study of Complex Systems at the University of Michigan for hospitality while some of this work was conducted.
The file contains Supplementary Figures 1-7 with Legends and Supplementary Tables 1-2.
Rod and cone photoreceptors detect light and relay this information through a multisynaptic pathway to the brain by means of retinal ganglion cells (RGCs)
The retinal ganglion cells that express melanopsin (rendering them intrinsically photosensitive) send monosynaptic projections to the suprachiasmatic nucleus (SCN) and the intergeniculate leaflet (IGL), responsible for circadian photoentrainment, and the olivary pretectal nucleus (OPN), responsible for PLR
To eliminate ipRGCs, we introduced a gene (aDTA) encoding attenuated diphtheria toxin A subunit (aDTA)
To ablate ipRGCs more completely, we generated animals homozygous for aDTA (Opn4
To assess whether image-forming functions are affected in animals expressing aDTA, we measured electroretinograms, optokinetic nystagmus responses, visual acuity and the ability of the animals to detect a visual cue. We found that the electroretinograms and optokinetic nystagmus responses were normal in animals lacking ipRGCs (
Pupil constriction regulates the amount of light entering the eye, and thus pupil diameter is negatively correlated with light intensity. At high light intensity, the iris decreases the area of the pupil by 95% (full constriction) in comparison with dark-adapted conditions (fully dilated). At low light intensity in which the pupil constricts by 50% or less, rod–cone input is the main signal
At high light intensity, six of nine Opn4
The aDTA homozygotes have greater ablation of ipRGCs, and consequently all 12 Opn4
To assess the contribution of ipRGCs to circadian photoentrainment, we analysed the wheel running activity of wild-type (n = 11) and Opn4
The acute effects of light on activity (also known as masking) can be studied by using an ultradian 7-h light/dark (3.5:3.5 LD) cycle that disrupts the oscillator
We have shown that the loss of ipRGCs does not influence image formation and therefore the involvement of this ganglion cell class in classical vision is only modulatory
All experiments were conducted in accordance with National Institutes of Health guidelines and were approved by institutional animal care and use committees of the universities involved.
We used behavioural tests measuring the integrity of the outer retina (electroretinograms)
Using the homologous arms that we used previously
Whole retinas from Opn4
Animals anaesthetized by intraperitoneal injection of Avertin (20 ml kg
Mice were anaesthetized with Ketamine (80 mg kg
Animals maintained under 12:12 light/dark cycles for three days were used to collect data within the light phase (after 50 min of dark adaptation). The electroretinogram set-up was similar to that used previously
White light (10 ms in duration) was provided by a
A mouse stabilized with a head post was placed into an acrylic holder in a 12-inch diameter drum. Computer-generated stimuli were projected down onto the drum walls. Black and white stripes 4° in width were rotated at 5° s
To assess the ability of mice to detect a visual cue, we trained the animals to find a platform marked by a 10-cm tall, high-contrast visual cue under bright light (500 lx) in an 85-cm pool. On day 1, mice were trained with four trials 15 min apart. On the following day, latency to find the island was recorded first with the cue and then without it.
All animals were dark-adapted for at least 1 h, and the eye of each animal receiving the photic stimulus was treated with 0.1% atropine before the start of recording. Measurements were restricted to the middle of the subjective day (CT4–8). One eye of each mouse was digitally captured at a frequency of one image per second for 63 s with a charge-coupled-device camera. The light stimuli (xenon arc light source) consisted of a 60-s pulse at an intensity of 3.8 mW cm
For the Opn4
Mice were placed in cages with a 4.5-inch running wheel, and their activity was monitored with
a, Model describing how rod–cone signalling through conventional RGCs or ipRGCs contributes to NIF functions. The role of ipRGCs in image formation is speculative (dotted line). b, Melanopsin antibody staining in retinas of 18-month-old wild-type (n = 6) and Opn4
a–c, X-gal staining in Opn4
a, All nine Opn4
We thank J. Mackes and G. Harrison for help in genotyping the animals; R. Kuruvilla, M. Van Doren, B. Wendland, M. Halpern, M. Caterina, C.-Y. Su, J. Bradley and laboratory members in the Biology Department at the Johns Hopkins University for scientific discussions and comments on the manuscript. This work was supported by grants from the National Institutes of Health (to S. Hattar and K.-W.Y.), the Biotechnology and Biological Sciences Research Council (to R.J.L.) and the David and Lucile Packard and Alfred P. Sloan Foundations (to S. Hattar).
Author Contributions A.D.G. and S. Hattar wrote the paper. J.L.E., R.J.L., D.M.B. and T.C.B. gave helpful comments on the manuscript. A.D.G., J.L.E. and C.M.A. in S. Hattar’s laboratory performed all the behavioural studies on the aDTA homozygous animals, as well as the X-gal staining of the Opn4
The file contains Supplementary Methods and Supplementary Figures 1-4 with Legends.
The aryl hydrocarbon receptor (AHR) is a ligand-dependent transcription factor best known for mediating the toxicity of dioxin
The AHR is a ligand-dependent transcription factor that mediates a range of critical cellular events in response to halogenated aromatic hydrocarbons and non-halogenated polycyclic aromatic hydrocarbons such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
Gene array analysis of CD4
AHR resides in the cytoplasm in complex with Hsp90 until binding of ligand triggers conformational changes resulting in an exchange of Hsp90 for the nuclear translocation component ARNT (reviewed in ref.
To test whether exposure of T cells to FICZ influences differentiation of naive CD4
Comparison of TH17 differentiation in CD4
To test whether Ahr expression on its own is essential and sufficient to drive IL-22 expression, we performed retroviral transduction of sorted naive AHR-deficient CD4
Next we immunized B6 mice with myelin oligodendrocyte peptide 35–55 (MOG35–55) in complete Freund’s adjuvant (CFA), removed draining lymph nodes 7 days later and isolated CCR6
TH17 T cells have a prominent role in the pathology of autoimmune diseases such as experimental autoimmune encephalomyelitis (EAE), which is induced by immunization with MOG peptide 35–55 and CFA. Analysis of spinal cord at the height of the EAE response on day 18 showed increased numbers of CD4
Whereas B6 mice developed EAE with a mean day of onset of 13.9, AHR-deficient mice developed EAE with delayed kinetics (mean day of onset 15.6) in line with the attenuated TH17 differentiation seen in vitro. Despite the delayed onset of EAE, most AHR-deficient mice succumbed to disease eventually (
Studies on the effect of AHR stimulation on the immune system have so far focused exclusively on TCDD as a ligand because of its toxicological relevance. Although adverse effects of TCDD on immune responses are well documented
Our data show that AHR, in addition to promoting the expression of IL-22, enhances TH17 cell development and the expression levels of IL-17A and IL-17F, and consequently increases autoimmune pathology. Blockade of IL-17A with an auto-vaccine as well as neutralizing antibody can completely prevent the development of EAE
It is currently thought that IL-22 expression depends on induction by IL-23 (refs
How AHR interacts with the TH17 pathway is currently unknown, but there is a substantial amount of literature (reviewed in ref.
Autoimmune diseases are multifactorial, depending on intrinsic components such as genetics, hormones or age, and environmental factors, including infections, diet, drugs and chemicals. The increasing prevalence of certain autoimmune diseases in highly industrialized countries is probably connected to such environmental factors.
Our data linking a transcription factor responsive to environmental pollutants to the TH17 programme open intriguing possibilities regarding the potential of such factors to initiate or augment autoimmune conditions, and warrant closer examination of a possible role of AHR in human autoimmune diseases.
C57BL/6 (B6) and AHR-deficient mice on a B6 background (B6 BRA AHRKO)
Human peripheral blood mononuclear cells from a healthy volunteer were isolated by Ficoll/Paque, and CD4
A detailed description of procedures is given in Methods.
The expression of mRNA for transcription factors and cytokines in CD4
AHR was cloned into vector
EAE was induced and scored as described previously
RNA was extracted using Trizol (Invitrogen) and
Mouse: Hprt, Mm00446968_m1; Tbx21, Mm00450960_m1; Gata3, Mm00484683_m1; Foxp3, Mm00475156_m1; Rorc, Mm01261019_g1; Ahr, Mm00478930_ml; Arnt, Mm00507836_ml; Ahr exon1 forward, 5′-CGCCTCCGGGACGCAGGTGG-3′; Ahr exon2 reverse, 5′-AAAGAAGCTCTTGGCCCTCAG-3′; Ifng, Mm00801778_m1; Il4, Mm00445259_m1; Il17a, Mm00439619_m1; Il17f, Mm00521423_m1; Il22, Mm00444241_m1; Cyp1a1, Mm00487217_m1; Il23r, Mm00519942_m1. Human: HPRT, HS99999909_m1; IL17A, HS00174383_m1; IL17F, HS00369400_m1; IL22, HS00220924_m1; AHR, HS00169233_m1; CYP1A1, HS00153120_ml.
Conditions for different cell subsets were: TH0 (anti-IFN-&ggr; plus anti-IL-4 plus anti-TGF-&bgr;), TH1 (4 ng ml
The AHR ligands
FACS-sorted naive CD4
5 × 10
a, Fluorescence-activated cell sorter (FACS)-sorted naive CD4
a, Mean mRNA levels ± s.d. from B6 CD4
a, b, FACS-sorted naive CD4
a, RT–PCR analysis for Il17a, Il17f, Il22, Rorc (ROR-&ggr;t) and Ahr in FACS-sorted CCR6
We thank A. Smith for providing us with the AHR-deficient mouse strain; A. Rae and G. Preece for cell sorting; and Biological Services at NIMR, especially T. Norton and H. Boyes, for animal care and EAE scoring. This work was funded by the Medical Research Council UK.
Author Contributions M.V. and K.H. performed the experiments; A.M.W. and J.B. did the microarrays and analysis; L.D. and J.-C.R. generated and provided the anti-IL-22 antibody; and B.S. directed the research and wrote the manuscript.
The file contains Supplementary Results; Supplementary Tables 1-7; Supplementary Figures 1-8 with Legends and additional references.
The Philadelphia chromosome, a chromosomal abnormality that encodes BCR–ABL1, is the defining lesion of chronic myelogenous leukaemia (CML) and a subset of acute lymphoblastic leukaemia (ALL)
Acute lymphoblastic leukaemia (ALL) comprises a heterogeneous group of disorders characterized by recurring chromosomal abnormalities including translocations, trisomies and deletions. An ALL subtype with especially poor prognosis is characterized by the presence of the Philadelphia chromosome arising from the t(9;22)(q34;q11.2) translocation, which encodes the constitutively activated BCR–ABL1 tyrosine kinase. BCR–ABL1-positive ALL constitutes 5% of paediatric B-progenitor ALL and approximately 40% of adult ALL
To extend this analysis and identify lesions that distinguish CML from BCR–ABL1 ALL, we have now examined DNA from leukaemic samples from 304 paediatric and adult cases of ALL (254 B-progenitor; 50 T-lineage), including 21 paediatric and 22 adult BCR–ABL1 ALL cases, and 23 adult CML cases (
Ikaros is a member of a family of zinc-finger nuclear proteins that is required for normal lymphoid development
The expression of aberrant, dominant-negative Ikaros isoforms in B- and T-lineage ALL has been previously reported by several groups
The high frequency of focal deletions in IKZF1 in BCR–ABL1 ALL suggests that expression of alternative IKZF1 transcripts may be the result of specific genetic lesions, and not alternative splicing of an intact gene. To explore further this possibility, we performed reverse-transcriptase PCR (RT–PCR) analysis for IKZF1 transcripts in 159 cases (
To identify copy number alterations in CML, we performed SNP array analysis on 23 CML cases. In addition to chronic-phase CML (CP-CML), we also examined matched accelerated phase (AP-CML, N = 7) and blast crisis (BC-CML, N = 15 (12 myeloid and 3 lymphoid)) samples (
To explore the mechanism responsible for the identified IKZF1 deletion, we sequenced the IKZF1 &Dgr;3–6 genomic breakpoints (
We have identified a high frequency of copy number alterations in BCR–ABL1 ALL and BC-CML, but not in CP-CML. We observed a near obligate deletion of IKZF1 in BCR–ABL1 ALL, with 83.7% of paediatric and adult cases containing deletions that lead to a reduction in dose and/or the expression of an altered Ikaros isoform. By contrast, deletion of IKZF1 was not detected in CP-CML, but was identified as an acquired lesion in two of three lymphoid BC-CML samples. These data, together with the low frequency of IKZF1 deletions in other paediatric B-progenitor ALL cases, and the lack of focal IKZF1 aberrations in recently reported genomic analysis of non-haematopoietic tumours
Two-hundred and eighty-two paediatric ALL cases, 22 adult BCR–ABL1 ALL cases, 49 samples obtained from 23 adult patients with chronic myeloid leukaemia (CML) and 36 leukaemia cell lines were studied (
Patients and samples comprised 282 patients with acute lymphoblastic leukaemia (ALL) treated at St Jude Children’s Research Hospital, 22 adult BCR–ABL1 ALL patients treated at the University of Chicago, and 49 samples obtained from 23 adult patients with chronic myeloid leukaemia (CML) treated at the Institute of Medical and Veterinary Science, Adelaide (
Thirty-six acute myeloid and lymphoid leukaemia cell lines were genotyped using the Affymetrix Mapping 250k Sty and Nsp arrays. These were the ALL cell lines 380 (MYC–IGH and BCL2–IGH B precursor), 697 (TCF3–PBX1), AT1 (ETV6–RUNX1), BV173 (CML in lymphoid blast crisis), CCRF-CEM (TAL–SIL), Jurkat (T-ALL), Kasumi-2 (TCF3–PBX1), MHH-CALL-2 (hyperdiploid B-precursor ALL), MHH-CALL-3 (TCF3–PBX1), MOLT3 (T-ALL), MOLT4 (T-ALL), NALM-6 (B-precursor ALL), OP1 (BCR–ABL1), Reh (ETV6–RUNX1), RS4;11 (MLL–AF4), SD1 (BCR–ABL1), SUP-B15 (BCR–ABL1), TOM-1 (BCR–ABL1), U-937 (PICALM–AF10), UOCB1 (TCF3–HLF), YT (NK leukaemia), and the AML cell lines CMK (FAB M7), HL-60 (FAB M2), K-562 (CML in myeloid blast crisis), Kasumi-1 (RUNX1–RUNX1T1), KG-1 (myelocytic leukaemia), ME-1 (CBFB–MYH11), ML-2 (MLL–AF6), M-07e (FAB M7), Mono Mac 6 (MLL–AF9), MV4-11 (MLL–AF4), NB4 (PML–RARA), NOMO-1 (MLL–AF9), PL21 (FAB M3), SKNO-1 (RUNX1–RUNX1T1) and THP-1 (FAB M5). Cell lines were obtained from the Deutsche Sammlung von Mikroorganismen und Zellkulturen, Braunschweig, Germany, the American Type Culture Collection, Manassas, Virginia, from local institutional repositories, or were gifts from O. Heidenreich (SKNO-1) and D. Campana (OP1). Cells were cultured in accordance with previously published recommendations
Collection and processing of diagnostic and remission bone marrow and peripheral blood samples for Affymetrix SNP microarray analysis has been previously reported in detail
FISH for IKZF1 deletion was performed using diagnostic bone marrow or peripheral blood leukaemic cells in Carnoy’s fixative as previously described
RNA was extracted and reverse transcribed using random hexamer primers and
Whole-cell lysates of 3–6 × 10
Methylation status of the IKZF1 promoter CpG island (PCR amplicon hg17 coordinates: chromosome 7 50121508–50121714) was performed using MALDI-TOF mass spectrometry of PCR-amplified, bisulphite-modified genomic DNA extracted from leukaemic cells as previously described
Associations between ALL subtype and IKZF1 deletion frequency were calculated using the exact likelihood ratio test. Differences in Ik6 expression between IKZF1 &Dgr;3–6 and non-&Dgr;3–6 cases were assessed using the exact Wilcoxon–Mann–Whitney test. All P-values reported are two-sided. Analyses were performed using
a, Domain structure of IKZF1. Exons 3–5 encode four N-terminal zinc fingers (black boxes) responsible for DNA binding. The C-terminal zinc fingers encoded by exon 7 are essential for homo- and heterodimerization. b, Genomic organization of IKZF1 and location of each of the 36 deletions observed in BCR–ABL1 B-progenitor ALL. Each line depicts the deletion(s) observed in each case. In four cases, two discontiguous deletions were observed. Hemizygous deletions are solid lines and homozygous deletions dashed. Arrows indicate deletions extending beyond the limits of the figure. The exact boundaries of the deletions were defined by genomic quantitative PCR, and for IKZF1 &Dgr;3–6, by long-range genomic PCR (red arrow). c, dChip SNP raw log2 ratio copy number data depicting IKZF1 deletions for 29 BCR–ABL1 cases and 3 B-progenitor ALL cell lines.
a, Domain structure of the IKZF1 isoforms detected by RT–PCR, examples of which are shown in b. b, RT–PCR for IKZF1 transcripts (using exon 0- and 7-specific primers) in representative cases with various IKZF1 genomic abnormalities. Each case expressing an aberrant isoform has a corresponding IKZF1 genomic deletion. IKZF1 &Dgr;3–6 was also detected in the BCR–ABL1 ALL cell lines SUP-B15 and OP1, and &Dgr;1–6 in the ALL cell line 380. c, Western blotting for Ikaros using a C-terminus-specific polyclonal antibody. Ik6 was only detectable in cases with IKZF1 &Dgr;3–6. The &Dgr;1–6 and &Dgr;2–6 deletions do not produce a detectable protein. In three cases with multiple focal hemizygous deletions involving different regions of IKZF1 (BCR-ABL-SNP-26, BCR-ABL-SNP-29 and BCR-ABL-SNP-31), no wild-type Ikaros was detectable by RT–PCR or western blotting, indicating that the deletions involve both copies of IKZF1 in each case.
a, dChip SNP log2 ratio copy number heatmaps of four CML cases showing acquisition of IKZF1 deletions at progression to blast crisis. b, c, Pherograms of IKZF1 exon 7 sequencing demonstrating acquisition of the coding nucleotide 1520C>A, amino acid Ser507X mutation at chronic-phase (b) and blast crisis (c) in case CML-7. As this case has a concomitant hemizygous IKZF1 deletion involving exon 7, the mutation appears to be homozygous.
The authors thank Z. Cai for technical help, K. Rakestraw and J. Armstrong for assistance with sequencing, R. Williams and C. Sherr for the provision of Arf null hematopoietic cells and BCR–ABL1 retroviral vectors, O. Heidenreich for providing the SKNO-1 cell line, and D. Campana for providing the OP1 cell line. This study was supported by the American Lebanese Syrian Associated Charities of St Jude Children’s Research Hospital. C.G.M. was supported by grants from the National Health and Medical Research Council (Australia), the Royal Australasian College of Physicians, and the Haematology Society of Australasia.
Author Contributions C.G.M. collected and extracted clinical samples, performed laboratory assays and analysed data. C.B.M., L.A.P., J.D. and I.R. performed laboratory assays. J.M. analysed SNP array data. D.W., T.P.H., M.M.L., C.-H.P., M.V.R. and S.A.S. collected clinical samples and data. C.G.M and J.R.D designed the study and wrote the manuscript, which was reviewed by all authors.
The primary SNP microarray data have been deposited in NCBIs Gene Expression Omnibus (GEO,
The file contains Supplementary Figures 1-14 with Legends. and Legends to Supplementary Movies 1-5.
The file contains Supplementary Table S1. This file contains a list of RNAs significantly enriched in pseudopodia in response to both LPA and fibronectin (FN) stimulation.
The file contains Supplementary Table S2. This file contains a list of RNAs significantly enriched in pseudopodia (Ps) or cell bodies (CB) in response to LPA.
The file contains Supplementary Table S3. This file contains a list of RNAs significantly enriched in pseudopodia (Ps) or cell bodies (CB) in response to Fibronectin.
The file contains Supplementary Movie 1. Confocal fluorescence time lapse imaging of a cell expressing mRFP (red), MS2-GFP (green) and the &bgr;globin-24bs/pkp4 mRNA. Overlay images of the two channels are shown. Arrows point to protrusions with localized RNA granules. Note that the RNA granules remain stationary over the course of observation. Time is indicated on the upper left corner in seconds.
The file contains Supplementary Movie 2. FRAP experiment of a localized RNA granule. Shown is a cell expressing MS2-GFP and the &bgr;globin-24bs/pkp4 mRNA. A localized RNA granule at the end of a protrusion is indicated by an arrow. Fluorescence at this protrusion was bleached and its recovery recorded over ca 60 seconds. Note that fluorescence recovers only minimally during the course of observation. Time is indicated on the upper left corner in seconds.
The file contains Supplementary Movie 3. FRAP experiment of a localized RNA granule. Shown is a cell expressing MS2-GFP and the &bgr;globin-24bs/pkp4 mRNA and which exhibits two localized RNA granules at the ends of two protrusions. Fluorescence at one of the granules (arrow) was bleached and its recovery recorded over ca 4 minutes. Note that fluorescence recovers only minimally during the course of observation while the second granule (arrowhead) remains stationary. Time is indicated on the upper left corner in seconds.
The file contains Supplementary Movie 4. FRAP experiment of a cell expressing EB1-GFP. Fluorescence at a protrusion (arrow) was bleached and its recovery recorded over ca 60 seconds. Note that EB1-GFP comets (representing +ends of dynamic MTs) rapidly move throughout the cell body and several of them enter into protrusions over the course of observation. Time is indicated on the upper left corner in seconds.
The file contains Supplementary Movie 5. FRAP experiment of a cell expressing GFP-APC which exhibits APC granules at the tips of different protrusions. Fluorescence at the tip of one protrusion (arrow) was bleached and its recovery recorded over ca 60 seconds. Note that fluorescence recovers only minimally during the course of observation while another APC granule (arrowhead) remains stationary. Time is indicated on the upper left corner in seconds.
RNA localization is important for the establishment and maintenance of polarity in multiple cell types. Localized RNAs are usually transported along microtubules or actin filaments
To identify on a genome-wide scale RNAs that are enriched at the leading edge of migrating cells, we used a fractionation method in which cells are plated on a microporous filter and then induced to polarize and extend pseudopodial protrusions in response to a migratory stimulus. Pseudopodia and cell bodies are then physically isolated and their contents compared
About 50 RNAs were significantly enriched in pseudopodia in response to both migratory stimuli (
To dissect the localization mechanisms, we focused on the Rab13 and plakophilin 4 (Pkp4) messenger RNAs, which both showed robust localization. First, we expressed in NIH/3T3 cells the Rab13 gene encompassing the whole open reading frame and 3′ untranslated region (UTR) (
To visualize the localization pattern conferred by the Rab13 3′ UTR, we used the MS2 system
Twenty-four repeats of the MS2-binding site were introduced into the &bgr;-globin gene downstream of the coding region, followed by different 3′ UTR sequences (
When imaged over time, the localized RNA granules are relatively stationary (
To gain insight into the identity of the structures that anchor transcripts at the tips of protrusions, the localized RNAs were expressed together with fluorescently tagged markers of various cellular structures. The RNA granules did not co-localize significantly with DCP1-containing P-bodies, focal adhesions (
The fact that the RNA granules remain stationary over several minutes (
APC is an unusual plus-end tracking protein that associates with only a minority of microtubules, and it has been observed in particular at the plus ends of Glu-microtubules
To test this hypothesis, we asked if these RNAs bind specifically to endogenous APC. Indeed, APC co-immunoprecipitated with both Rab13 and Pkp4 mRNAs (
To test directly whether APC mediates anchoring of the localized RNAs, we knocked down APC expression using short-hairpin RNAs (shRNAs) (
This study provides the first genome-wide identification of asymmetrically distributed RNAs in fibroblasts. We show that, in response to migratory stimuli, >50 mRNAs accumulate in cellular protrusions of fibroblasts, revealing that even in less differentiated polarized cells, RNA localization mechanisms are widely used. Several RNAs have also been recently reported to associate in vitro with mitotic microtubules, but the mechanism is unknown
To isolate pseudopodia and cell bodies in response to LPA, we followed the protocol described by ref.
RNA was isolated using
Biotin-labelled cRNAs were generated from total RNA of pseudopodia and cell body fractions and were hybridized to
A plasmid encoding the human &bgr;-globin mRNA was provided by J. Lykke-Andersen. The plasmid contains all the coding exons and introns of human &bgr;-globin, 54 nucleotides of the &bgr;-globin 5′ UTR, 69 nucleotides of the &bgr;-globin 3′ UTR followed by 24 repeats of the MS2-binding site cloned into the HindIII, XbaI sites of
The plasmid pcNMS2, containing an oligomerization-defective MS2 coat protein mutant, was provided by J. Lykke-Andersen. The GFP sequence was PCR amplified from plasmid
For knockdown experiments, oligonucleotides were synthesized targeting different regions of the mouse Apc mRNA. Sequences of the oligonucleotides are as follows: shAPC 2 sense oligonucleotide, 5′-GATCCCCGAATCAACCAGGCATAATA
NIH/3T3 cells were grown in DMEM supplemented with 10% calf serum, sodium pyruvate, penicillin and streptomycin (Invitrogen). For live cell imaging, plasmid constructs were transfected with
To isolate pseudopodia and cell bodies of cells induced to migrate with LPA, we followed the protocol described by ref.
For fractionation after nocodazole treatment, cells were processed as above except that nocodazole was added to 10 &mgr;M during the last 25 min of the assay.
To isolate pseudopodia and cell bodies of cells migrating towards fibronectin, ∼1 × 10
Total RNA or RNA from pseudopodia and cell body fractions was isolated using
The identity of the amplified products was verified either by sequencing or in the case of transfected RNAs by ensuring that no product was amplified when using RNA from vector-transfected cells. PCR reactions contained 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.2 mM each dNTPs, 0.5 &mgr;M each primers and 2.5 U Taq polymerase. The amount of cDNA and the number of cycles were varied for each primer pair, to ensure amplification was within the linear phase. This was verified in all experiments by including decreasing amounts of selected samples.
To calculate the enrichment of RNAs in pseudopodia, equal amounts of RNA from pseudopodia and cell body fractions were analysed by RT–PCR, the signals were quantified using WCIF ImageJ software, normalized to the control Arpc3 mRNA and enrichment in pseudopodia was defined as (pseudopodia signal/cell body signal)-1, when pseudopodia signal > cell body signal or as (-1)×[(cell body signal/pseudopodia signal)-1], when pseudopodia signal < cell body signal.
Microarray analysis was performed at the Biomolecular Research Facility at the University of Virginia. Total RNA from pseudopodia and cell body fractions was analysed on an
In all hybridization experiments, quality assessment variables, such as background, noise, GAPDH 3′/5′ ratio, were within the acceptance limits.
For experiments analysing pseudopodia and cell body fractions from cells induced with LPA, results were analysed using the GeneChip Operating Software (GCOS) platform. RNAs were considered to be significantly enriched in pseudopodia if the P-value was significant, the fold change in signal intensity was greater than 2.2 and the absolute difference in signal intensity was greater than 100.
For experiments analysing pseudopodia and cell body fractions from cells migrating towards fibronectin, results were analysed using the dChip software (
For western blot and immunofluorescence, the following antibodies were used:
For immunoprecipitation, anti-APC (C-20) or control antibody (
Imaging and photobleaching were performed with a
a, Schematic diagram depicting strategies used for isolation of pseudopodia (Ps) and cell bodies (CB). NIH/3T3 cells, plated on microporous filters, were induced to extend protrusions by adding LPA in the bottom chamber. Alternatively, the underside of microporous filters was coated with fibronectin (FN) and cells plated on top extended protrusions towards the fibronectin-coated surface. Cell body and pseudopodia fractions were subsequently isolated. b, Cells extending protrusions in response to LPA or fibronectin, as described in a, were stained with fluorescein-isothiocyanate-conjugated phalloidin. Confocal images of the top and bottom side of the filter are shown. Scale bar, 15 &mgr;m. c, Total RNA from pseudopodia and cell body fractions of cells extending protrusions in response to LPA was analysed by RT–PCR to detect the mRNAs indicated on the left. Increasing amounts of the pseudopodia sample were amplified (lanes 1 and 2) to ensure linearity of the amplification. Values on the right indicate mean pseudopodia/cell body ratios, normalized to the control Arpc3 mRNA, ± s.e.m., n = 3.
a, Schematic of transfected constructs. Black boxes, exons; black lines, introns; grey boxes, 3′ UTR; white box, Flag tag. b, Cells transfected with the constructs depicted in a were fractionated into pseudopodia and cell body fractions after induction with LPA, and RNA enrichment towards each fraction was calculated (n = 2–4). Error bars, s.e.m. c, Schematic depicting the general structure of &bgr;-globin constructs used in d and e. 24bs, 24 MS2-binding sites. d, Imaging of live cells co-transfected with plasmids encoding mRFP, MS2–GFP and &bgr;-globin constructs with various 3′ UTRs. Shown are representative examples of the localization patterns observed when &bgr;-globin mRNA carried the UTRs indicated on the left (localized in granules at tips of protrusions (arrows) or diffuse in the cytoplasm). Scale bar, 10 &mgr;m. e, Quantification of the percentage of cells exhibiting localized RNA distribution when transfected as in d, with constructs carrying the indicated 3′ UTRs (n = 3). Error bars, s.d.
a, Cells were co-transfected with MS2–GFP and the &bgr;-globin-24bs/Pkp4 RNA. Fluorescence intensity in granules at protrusive areas (left panel, red circle) or within the cytoplasm (right panel, red circle) was monitored before and after photobleaching. The arrow indicates time of bleach. Curves represent average values of ten and five independent experiments, respectively. Error bars indicate s.d. b, c, Confocal fluorescence images of cells expressing MS2–GFP, the &bgr;-globin-24bs/Pkp4 RNA and either RFP–tubulin (b) or fixed and stained with anti-Glu-tubulin antibody (c). Panels show edges of protrusive areas. Scale bar, 3 &mgr;m. d, Cells expressing mRFP, MS2–GFP and the &bgr;-globin-24bs/Pkp4 RNA were treated with nocodazole (Noc.) or cytochalasin D (Cyt.D). The percentage of cells exhibiting localized RNA distribution was quantified as in
a, Confocal fluorescence images of cells expressing MS2–GFP, the &bgr;-globin-24bs/Pkp4 RNA and 3×OFP–APC. Panels show edges of protrusive areas. b, NIH/3T3 lysates were immunoprecipitated (IP) with control antibody (IgG), anti-APC antibody (anti-APC) or no antibody (-) and analysed to detect the indicated proteins and mRNAs. c, Same as in b, except that before immunoprecipitation lysates were treated (+) or not (-) with RNase. d, Lysates of cells transfected with Flag–FMRP or vector were immunoprecipitated with anti-Flag antibody and analysed to detect the indicated proteins and mRNAs. e, Confocal fluorescence images of cells expressing MS2–GFP, the &bgr;-globin-24bs/Pkp4 RNA and RFP–FMRP. Panels show edges of protrusive areas. f, Cells were transfected with shRNAs against luciferase (shLuc) or Apc (shAPC), individually or in combination as indicated. Left panel, normalized Apc mRNA levels; right panel, enrichment of the indicated mRNAs in pseudopodia in response to LPA (n = 2–3); asterisk, P-value < 0.05 by paired, two-tailed t-test versus shLuc control. g, shLuc- and shAPC-expressing cells were transfected with MS2–GFP and the &bgr;-globin-24bs/Pkp4 RNA. The percentage of cells exhibiting localized RNA distribution, as described in
We thank Y. Bao for the microarray and real-time PCR analysis, and J. Lykke-Andersen, I. Nathke, J. T. Parsons, R. Darnell and G. Gundersen for plasmids and antibodies. S.M. is a fellow of the Leukemia and Lymphoma Society. This work was supported by a grant from the NIH to I.G.M., and by the James and Rebecca Craig Foundation.
Author Contributions S.M. performed the experiments. S.M., K.M. and I.G.M. designed the experiments and analysed the data. K.M. provided reagents. S.M. and I.G.M. wrote the manuscript.
The microarray data have been deposited in NCBI's Gene Expression Omnibus (GEO,
Life without RNase P p.120version="1.0" encoding="utf-8"?> ]>
The file contains Supplementary Figures S1-S2 wit Legends.
The universality of ribonuclease P (RNase P), the ribonucleoprotein essential for transfer RNA (tRNA) 5′ maturation
The transcription of tRNA genes generates precursor tRNA molecules with extended 5′ and 3′ termini that are processed into mature tRNA
Surprisingly, analysis of genome sequence failed to reveal genes for RNase P in N. equitans, several species of Pyrobaculum and the bacterium Aquifex aeolicus
First we used a computational approach to verify that no RNase P RNA is present in the genome of N. equitans. An algorithm was developed which, based on the observation that non-coding RNA genes are easily identified in AT-rich hyperthermophiles
The lack of RNase P activity raised the question of how an organism could live without RNase P and led us to investigate whether tRNAs are produced without 5′ leader sequences. This hypothesis would require that transcription of all tRNA species starts at the 5′ terminus of the mature tRNA. Therefore we analysed the promoter placement upstream of all tRNA (and tRNA half
This strict promoter placement probably allows production of leaderless transcripts. Transcription is predominantly initiated with a purine (G > A) in all living organisms
Transcription initiation at the mature 5′ termini of nanoarchaeal tRNAs would result in triphosphorylated 5′ termini (for example found in the initiator tRNA of Halobacterium volcanii
The presence of a catalytic RNA moiety and the universal conservation of RNase P imply that this enzyme is ancient, possibly a remnant of the RNA world
Because tRNA is the typical substrate of RNase P, it is possible that tRNA and tRNA genes predate RNase P. The emergence of RNase P would then coincide with a more flexible tRNA promoter placement or tRNA gene duplications. Genome reorganization and horizontal gene transfer are sometimes associated with recombination at tRNA genes. Such genomic perturbations allow an increased evolutionary tempo. Thus RNase P may have coevolved with an increased rate of horizontal gene transfer that facilitated the evolution of the genetic code.
Oligonucleotide synthesis and DNA sequencing was performed by the Keck Foundation Biotechnology Resource Laboratory at Yale University. N. equitans cells were obtained from K. O. Stetter and M. Thomm.
The N. equitans initiator pre-tRNA gene was cloned with the leader sequence 5′-GGUUAUAACUUACU-3′ into a pUC19 vector that allowed for in vitro T7 RNA polymerase run-off transcription after plasmid cleavage with NsiI. The tRNA was internally labelled in the presence of [&agr;-
Five micrograms of the total RNA of M. kandleri, N. equitans and P. aerophilum were incubated with 5 U of
P. aerophilum and N. equitans cells were each re-suspended in buffer containing 50 mM Tris·HCl (pH 7.5), 500 mM NaCl and 3 mM DTT, broken by sonication and subsequently centrifuged for 30 min at 30,000g. RNase P cleavage assays were performed with radioactively labelled tRNA transcript for 30 min at 37 °C in buffer containing 50 mM HEPES (pH 7.5), 20 mM MgCl2 and 500 mM NH4Cl, either using E. coli RNase P RNA transcript or cell extracts. Cleavage assays with archaeal cell extracts were performed with variation of the incubation temperature from 37 °C to 80 °C, incubation time from 10 min to 3 h, and with variation of salt concentration from 10 to 100 mM MgCl2 and from 100 to 1,500 mM NH4Cl.
The genes for N. equitans tRNAi
The N. equitans metS gene (NEQ457) and tyrs gene (NEQ389) were amplified by PCR from genomic DNA. The metS gene was cloned into the XhoI/BlpI site of
Purification of total small RNA from P. aerophilum and N. equitans cells, circularization, reverse transcription with
a, RNase P cleavage assay. The internally radioactively labelled pre-tRNAi
The upstream region of all 44 tRNA genes (including 5′ tRNA halves) was aligned. Position 1 indicates the first tRNA nucleotide. Position -1 of three tRNA species is boxed in red. The conserved promoter element is visualized by a WebLogo
Equal amounts of total small RNA from M. kandleri (Mk), N. equitans (Ne), P. aerophilum (Pa) and an N. equitans tRNA
The pre-tRNA leader sequences are indicated in red; RNase P proteins are indicated in blue.
We thank P. O’Donoghue, J. Yuan and L. Sherrer for help and encouragement. This work was supported by grants from the National Institute of General Medical Sciences and the Department of Energy (D.S.) and the National Science Foundation (I.S.).
The file contains Supplementary Tables 1-2 and Supplementary Figures 1-13 with Legends.
During infection by Gram-negative pathogenic bacteria, the type III secretion system (T3SS) is assembled to allow for the direct transmission of bacterial virulence effectors into the host cell
The EscU family of proteins is an essential component of the inner membrane ring of T3SS systems. It is thought to be composed of approximately half a dozen highly conserved proteins (EscJNRSTUV in enteropathogenic E. coli). Earlier primary sequence analysis of the EscU and orthologous FlhB (flagellar) family of proteins predicted two major domains: an approximate 200-amino-acid amino (N)-terminal domain (NTD) that is involved in the association to the T3SS inner membrane basal ring; and an approximate 100-amino-acid carboxy (C)-terminal domain (CTD), which is known to undergo auto-cleavage and to mediate a switch in the secretion substrate profile of the T3SS
The crystallographic data collectively show that the CTD domains are highly conserved in structure (despite the modest approximate 30% sequence identity of the region) (
EscU contains a cleavage site conserved within all T3SS and flagellar orthologues
To address the mechanism of cleavage further, we analysed six point mutations at the auto-cleavage site in EscU (N262A, N262D in two crystal forms (that is, N262D and N262Di), P263A, T264A, H265A and R313T) (
To assess the importance of the conserved residues (
In addition to electrostatic features (
EscU and orthologues in other T3SSs, as well as the flagellar system, have been implicated in acting as part of a molecular ‘switch’ that regulates the chronological secretion of apparatus components and virulence effectors (detailed in
Our structural and biochemical data point to the conserved linker region of EscU orthologues for playing a potential role in the switching events. The proposed linker region (EscU 215–245 and SpaS 210–241) connects the N-terminal membrane-anchored domain with the globular CTD. Although earlier sequence analysis predicted a helical linker
To analyse the importance of the observed linker in T3SS function, we used our structure to design three deletion mutations (EscU(-5) 232–236, EscU(-11) 234–245, EscU(-16) 230–245) and two point mutations (G229P and G235P) in the flexible hinge region of EscU. These mutants were screened in vivo for secretion of the filament component EspA, the translocon components (EspB/EspD) and for the effector Tir in &Dgr;escU and in wild-type (WT) backgrounds (
Protein purification, site-directed mutagenesis and secretion assays were performed as previously published
For purification of His-tagged SpaS&Dgr;210, EscU&Dgr;214 and all mutants, the expression construct pEscU/pSpaS was generated by PCR-amplifying the escU/spaS open reading frame from EPEC E2348/69 and S. typhimurium genomic DNA, and cloning it into the NheI/BamHI and NheI/HindIII sites of the
EscU mutants were generated using QuikChange site-directed mutagenesis (coding and antisense primers containing a single mutagenic site were used for PCR amplification).
Purified EscU&Dgr;214/SpaS&Dgr;210 (at 15–20 mg ml
The sacB gene-based allelic exchange method
Primers EPescU-HAF (SacI) (5′-C
EPEC strains were grown in Luria-Bertani and then subcultured in Dulbecco’s Modified Eagle Media to induce type III secretion as previously described
The following genetic manipulations were performed using the &lgr;-RED system
Secreted proteins were prepared as previously described for S. typhimurium
Overnight cultures of WT S. typhimurium and the spaS mutants were grown in Luria-Bertani medium at 37 °C. Cultures were passed six times through a hypodermic needle (
The circular dichroism spectra were recorded with a
a, The native cleaved CTD of EscU and SpaS with a blue arrow pointing to the auto-cleavage site. CTD is a novel &agr;/&bgr;-fold with a mixed parallel and anti-parallel five-stranded twisted &bgr;-sheet (topology &bgr;4, &bgr;1, &bgr;2, &bgr;3, &bgr;5) flanked by two helices on each side (&agr;1, &agr;2, &agr;3 and &agr;4). b, Superposition of the CTD reveals a different fold for the N-terminal linker between EscU, EscU mutants and SpaS, as well as a longer C-terminal helix for SpaS.
a, EPEC, escU mutants and complementation of type III secretion. Secretion of effector (Tir) and translocator proteins (EspB/D) checked under complementation of escU by WT and mutants. The cleavage state of EscU in vivo is presented for all EscU components. b, spaS WT, mutants and type III secretion. Secretions of effector (SipA), molecular ruler (InvJ), needle (PrgI) and translocator proteins (SipB/C) were analysed in WT and spaS mutants. The N258A and P259A substitutions in SpaS correspond to N262A and P263A respectively in EscU. Full-length blots/gels are presented in
a, Non-cleaved type I &bgr;-bend of P263A (left), the non-cleaved type II beta bend of N262A (middle left), the non-cleaved type II beta bend of N262Di (middle right) and the native cleaved loop with a flipped His 265 (right). Note the identical conformations of the N, C, C&agr; and C&bgr; at position 262 in all uncleaved forms. All maps are sigma-A weighted 2Fo - Fc electron density (1.5&sgr;); water molecules are represented by the red spheres. b, Detailed mechanism for the asparagine cyclization in EscU.
We thank T.S., M.C. and P.I.L. for discussions; the staff at the Advanced Light Source beamline 8.2.2 and SSRL beamline 11-1 for data collection time and assistance; and E. Galyov at the Institute for Animal Health for the SipA, SipB and SipC antibodies. This work was supported by an Izaak Walton Killam Research post-doctoral fellowship, a Michael Smith Foundation for Health Research (MSFHR) post-doctoral fellowship and a Canadian Institutes of Health Research (CIHR) post-doctoral fellowship (all to R.Z.). N.C.J.S. and B.B.F. thank the Howard Hughes International Scholar program and the CIHR for funding. N.C.J.S. also thanks the MSFHR and the Canada Foundation of Innovation for infrastructure funding support. N.C.J.S. is also an MSFHR Senior Scholar and CIHR Investigator. S.M. acknowledges grants from the National Institutes of Allergy and Infectious Diseases (NIAID), U54 AI057141 and 5RO1 AI030479.
Author Contributions R.Z. and M.V. cloned and purified EscU and SpaS. R.Z. crystallized EscU and SpaS. R.Z. solved the structures. H.B.F. isolated the chromosomal SpaS mutants. W.D., H.B.F and H.V.N. performed the biochemical experiments in
The atomic coordinates of all the structures have been deposited in the Protein Data Bank with the accession codes